Mathieu Lecarme wrote:
[EMAIL PROTECTED] a écrit :
And how could one create automatically such a summary?
Here's a site with some pointers to the literature and some systems out there to do summarization: http://www.summarization.com/ This is actually whole-document or even multiple-document summarization. Snippet production's a rather different problem, which needs to be sensitive to the query. What to show isn't so easy when there are many instances of the query terms in the document and very limited space.
Have a look to http://alias-i.com/lingpipe/index.html
> or http://www.nzdl.org/Kea/ We haven't written any kind of text summarization package for LingPipe. Kea works at a keyphrase level, not a doc summary level, though they reference a paper on using it for summarization: http://www.hicss.hawaii.edu/HICSS_35/HICSSpapers/PDFdocuments/DDUAC04.pdf Both LingPipe and Kea are able to find significant phrases, which is useful for query refinement or summarizing sets of search results, but not so useful for individual documents. It can be a huge help to add part-of-speech information to these kinds of approaches. - Bob Carpenter Alias-i --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
